Using advanced sentiment analysis to surface nuanced confirmations and/or contradictions from the call.
According to established social psychology research, the tone of the voice accounts for nearly 40% of an entire spoken message. Understanding how something is said when one is speaking can be almost as important as the textual content itself. As the industry-specific research below showcases, there are nuances to be detected utilizing tonal sentiment with textual sentiment across presentations, earnings calls and other verbal communications that present key inputs for different types of trading strategies.
Blackrock (BLK) is the world’s largest asset management company with over $9 trillion AUM in equities, fixed income, cash management, alternative investments, and real estate. It is also the largest provider of ETFs in the U.S. BLK stock closed at an all-time high of about $916 on July 12, 2021. The company had previously recorded positive earnings surprises in the four prior quarters, with an average beat of 11.1%. On July 14, 2021, Blackrock reported Q2 revenue growth of 32% and an earnings beat of 6%.
The BLK conference call that day featured the following participants:
This case study highlights the combination of textual sentiment, delivered via Aiera’s intelligence platform, and Helios’ own tonal sentiment scores derived from voice tone alone. Together, this textual and tonal information yields even better insights into the state of an organization.
Helios measures voice tone on a scale of -5 (most uncertain/uncomfortable) to 5 (most confident/bullish) with 0 as a neutral tone. The overall voice measurement on BLK’s Q2 call was 1.62 vs. their prior Q1 call of 2.44. Quarterly comparisons are important to understand as well as who was in attendance on these calls. Both of these calls are in the positive domain representing consistency with what they’ve said.
Throughout the call, CFO Gary Shedlin was remarkably consistent in his voice tone. The voice metrics of the three questions he answered were -0.07, -0.05, and -0.08, an extremely tight range around neutral. This confirms that his narrating voice tones are in fact quite similar to his answers on those particular questions. His answers to analyst questions were lengthy and covered a wide range of topics - comp rate, expenses, organic growth, current business cycle, client growth, and growth in Aladdin portfolio management tools.
CEO Larry Fink’s voice tone in his Q&A responses had a much wider range from a low of -5.0 to a high of 0.64. His most confident answer (0.64) was in response to a question from JP Morgan in which he addressed current and future growth of Blackrock’s LifePath and LifePath Paycheck retirement products. Within this response, he suggested that future announcements about LifePath are forthcoming, and the higher confidence metric here could be supportive of those announcements being positive or of his intimate knowledge of the LifePath portfolio.
Fink’s most uncertain answer (-5.0) was in response to Morgan Stanley’s question about Blackrock’s Aladdin portfolio management software, which had seen slowing customer implementations in the past. This was the last question of the conference call and he was providing some follow-up information after a more lengthy response from CFO Shedlin. Fink spoke bullishly about growing investor demand for data/analytics and concern about climate change being drivers for the Aladdin risk analytics products. The text sentiment scoring of his words was neutral. However, one could argue that the lack of conviction in his voice, as measured by Helios, suggests that this may be a domain that Fink is not intimately familiar with. For anchoring context, the closing remarks by CEO Fink rank near the highest on text sentiment given the bullish statements.
It was important to see that Shedlin’s response to the last question was also his least confident of the call (-0.08), although that means much less in the context of how relaxed he was overall. Moreover, we see text sentiment scoring in the negative on part two of the ACV response from CFO Shedlin. With a call that is heavily positive for text sentiment scoring, this negative average for a section that addresses the accelerated ACV is of high interest.
Fink’s second least confident answer (-1.24) was in response to Autonomous Research in which he was asked about any changes in investor behavior in anticipation of higher capital gains tax rates. Fink responded that investors seem to be more aware of the tax issue but their behaviors haven’t changed yet. The somewhat higher voice uncertainty might suggest customer attitudes towards capital gains tax rates are rapidly changing and an area of concern for investors to keep an eye on.
Overall, Blackrock’s earnings call showed a relatively high level of confidence in the voices of both CEO and CFO, however Helios’ data was able to pinpoint two specific topics that appeared to have increased emotional uncertainty (Aladdin portfolio management software and autonomous research), and one that signaled possible positive announcements (Lifepath retirement products), all worth further investigation by investors.
The textual sentiment, provided by AIera’s platform, coupled with the unique tonal sentiment from Helios unlocks a brand new channel for understanding corporate communications.
A recording of the earnings call can be found here, on Aiera’s platform, for full listening playback. Additional inquiries can be forwarded to firstname.lastname@example.org